RE: [OT] RE: Using URIs to identify non-information resources

Hi DanBri!  Thanks for your comments!  Responses in line below . . . .

> From: Dan Brickley [mailto:danbri@w3.org] 
> . . .
> Yep. I'm also very uncomfortable with the idea of 
> bottlenecking the identification of all non-"Information 
> Resource" entities through a handful of HTTP-based services.
> 
> To be very clear here, we are talking about *billions* of 
> objects. Look at the marketplace for RFID-based services.

Scalability (or efficiency) is precisely one reason I'm suggesting the
thing-described-by.org approach.  (The other is convenience.)  If you
instead take the custom 303-redirect approach, and everybody redirects
their own URIs, then those 303-redirects cannot be optimized away.
However, if you use URIs that are based on one or more well-known
303-redirect services (such as thing-described-by.org), then those extra
network accesses can be optimized away by inspecting the URI.  

The point is that if a service like thing-described-by.org is well
known, then there is no need to access it at all!  This seems to me to
be far more efficient and scalable (less of a processing bottleneck)
than the custom 303-redirect approach.

> That assumes knowledge of the 'http://thing-describe-by.org' 
> rules in the first place. 

Correct.

> And if thing-described-by.org is 
> not to be alone in the world, we'll need some way of finding 
> out other equivalent services. I'm skeptical.

They could be registered in Ontaria ( http://www.w3.org/2004/ontaria/ ),
they could be found by search engines, you could find out from your
friendly SemWeb neighbors.  All the same arguments about finding
ontologies apply to 303-redirect services like thing-described-by.org.
It's really the same situation.

> . . . How long do we expect 
> the service to be maintained for? (10 years? 20? 50?). What 
> procedures are in place to ensure the domain name fees are 
> paid on time during that period? . . . .

Those are legitimate concerns, but they are exactly the same concerns
that apply to PURL servers . . . only substantially less so, which I'll
explain.  

The big difference is that PURL servers *must* remain trustworthy
because they must continue to forward URLs to the appropriate place.
However, if a 303-redirect service like thing-described-by.org becomes
well known then its role as a forwarding service becomes completely
moot: there is no need to access it to know where a URI should forward.
Hence the service *cannot* deceptively forward you to the wrong place
without your knowledge.  Thus, once it becomes well known, the only
significant thing it does is to provide a formal delegation of
authority, such as
http://thing-described-by.org/#Delegation_of_Authority ,  to the
forwarded location.  But even that becomes moot when the service becomes
well known, because ultimately, that's what the users will expect it to
mean, even the statement on that page changes to say something
different.

This is analogous to the situation that could occur if the owners of a
well-known ontology try to change the meaning of
http://colors.example.org/blue to mean the color red, only much more
benign, because the only semantic that thing-described-by.org is
providing is a simple delegation of authority, which is trivial.

> I could go on, ... but I should make clear I'm not beating up 
> on the basic idea. It's a nice enough approach, if 
> implimented in a decentralised fashion. Implimented as a 
> Web-wide service, it would need a *lot* more policy and 
> support structures in place before it could be something I'd 
> encourage people to use.

Are you referring to the institutional backing that's needed for the
domain, mentioned earlier?  If not, what more do you think it would
need?

Thanks,
David Booth

Received on Saturday, 20 August 2005 02:12:19 UTC